Orbis-Cascade-Alliance / harvester

XForms-based OAI-PMH harvester for Orbis Cascade. Metadata are transformed into RDF and posted into a triplestore for access from finding aids.
9 stars 1 forks source link

View records after import #25

Closed adehner closed 7 years ago

adehner commented 7 years ago

In the beta harvester interface, it is possible to delete an imported set. But it isn't possible to view the records of "imported" sets.

I don't see that this is covered in the harvester specifications. But can this be enabled by adding link(s) for each set title on this page? If the user chose to contribute to both DPLA and Primo, ideally he/she would be able to view records transformed and imported for Primo and those for DPLA.

@jallibunn

ewg118 commented 7 years ago

Finally done. There will now be a link to view the records of the set from the final page of the harvester-interface. It displays 100 records per page as DPLA MAP RDF -> HTML. There is a simple pagination method to allow moving forward and backward through the pages.

ewg118 commented 7 years ago

By the way, the HTML view could be considered the groundwork for a browse interface. So far, it will either display all records in chronological order (most recent at the top) or accept a URL-encoded OAI-PMH set URL. See http://harvester-dev.orbiscascade.org/results?set=http%3A%2F%2Fcontent.libraries.wsu.edu%2Foai%2Foai.php%3Fverb%3DListRecords%26metadataPrefix%3Doai_dc%26set%3Dmatsura

Eventually, this can be expanded to yield results for faceted searches or keywords.

adehner commented 7 years ago

This looks good, and it's very cool that the groundwork is there for a browse interface!

But, we would like the "view records" link to display for each harvested set on this page ( https://harvester-dev.orbiscascade.org:8443/orbeon/harvester/admin/) so members can view their records at any time post-harvest.

ewg118 commented 7 years ago

Okay, this has been implemented.

adehner commented 7 years ago

Please update link text from "view this set as rdf" to...

if set was contributed to DPLA: "view records for DPLA" if set was contributed to Primo: "view records for Primo"

if set was contributed to both, then both links should appear

ewg118 commented 7 years ago

Done.

adehner commented 7 years ago

This morning I noticed a few weird things when clicking the "view records for Primo" link:

  1. When I click "view records for Primo" for the following sets (all from digital commons dams), I see a blank page instead of records:

    • All audio stories
    • Urban Studies and Planning Dissertations and Theses
    • Klipsun Magazine
  2. dc:spatial is blank in this set: Women in Sport at Western

  3. dc:rights is a concatenation of standardized and free-text rights statement in this set: Lee Moorhouse (1850-1926) photographs, 1888-1916

ewg118 commented 7 years ago

I think I have fixed the issue with 1. With 2., the dcterms:spatial itself is blank, so it's not a rendering problem, but a harvest problem. I checked the XSLT, and empty spatial/coverage values were not being filtered out in cases where there is a semicolon. 3. dc:rights is being split on semicolon again, but I thought we agreed to disable that since rights statements might contain them? I have forgotten now.

adehner commented 7 years ago

I'm still seeing instances of empty spatial/coverage for the walc set:

screen shot 2017-04-28 at 8 55 57 am
adehner commented 7 years ago

You are correct about that semicolon in rights! Thanks for reminding us that we don't want to split rights on semicolons. We ask that folks output standardized and free-text rights statements in separate properties, so the moorehouse set is a case for local cleanup.

ewg118 commented 7 years ago

I restored the dc:rights template so it doesn't split on semicolon.

On Fri, Apr 28, 2017 at 12:17 PM, adehner notifications@github.com wrote:

You are correct about that semicolon in rights! Thanks for reminding us that we don't want to split rights on semicolons. We ask that folks output standardized and free-text rights statements in separate properties, so the moorehouse set is a case for local cleanup.

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/Orbis-Cascade-Alliance/harvester/issues/25#issuecomment-298041785, or mute the thread https://github.com/notifications/unsubscribe-auth/AAiX8zbYg4zDjBCkZGVQ7Xxf8gXXNWviks5r0hE0gaJpZM4MkHAw .